I will be analyzing the famous iris data set!

The Data

The Iris flower data set or Fisher's Iris data set is a multivariate data set introduced by Sir Ronald Fisher in the 1936 as an example of discriminant analysis.

The data set consists of 50 samples from each of three species of Iris (Iris setosa, Iris virginica and Iris versicolor), so 150 total samples. Four features were measured from each sample: the length and the width of the sepals and petals, in centimeters.

Here's a picture of the three different Iris types:

In [1]:
# The Iris Setosa
from IPython.display import Image
url = 'http://upload.wikimedia.org/wikipedia/commons/5/56/Kosaciec_szczecinkowaty_Iris_setosa.jpg'
Image(url,width=300, height=300)
Out[1]:
In [2]:
# The Iris Versicolor
from IPython.display import Image
url = 'http://upload.wikimedia.org/wikipedia/commons/4/41/Iris_versicolor_3.jpg'
Image(url,width=300, height=300)
Out[2]:
In [3]:
# The Iris Virginica
from IPython.display import Image
url = 'http://upload.wikimedia.org/wikipedia/commons/9/9f/Iris_virginica.jpg'
Image(url,width=300, height=300)
Out[3]:

The iris dataset contains measurements for 150 iris flowers from three different species.

The three classes in the Iris dataset:

Iris-setosa (n=50)
Iris-versicolor (n=50)
Iris-virginica (n=50)

The four features of the Iris dataset:

sepal length in cm
sepal width in cm
petal length in cm
petal width in cm

Get the data

In [8]:
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt 
import seaborn as sns
%matplotlib inline
In [9]:
iris = sns.load_dataset('iris')

Exploratory Data Analysis

Creating a pairplot of the data set

In [11]:
sns.pairplot(iris, hue = 'species', palette = 'Dark2')
Out[11]:
<seaborn.axisgrid.PairGrid at 0x23ce24c31d0>

Creating a kde plot of sepal_length versus sepal width for setosa species of flower.

In [14]:
setosa = iris[iris['species']=='setosa'] 
sns.kdeplot(setosa['sepal_width'], setosa['sepal_length'], cmap ='plasma', shade = True, shade_lowest = False)
Out[14]:
<matplotlib.axes._subplots.AxesSubplot at 0x23ce31892e8>
In [30]:
iris.head()
Out[30]:
sepal_length sepal_width petal_length petal_width species
0 5.1 3.5 1.4 0.2 setosa
1 4.9 3.0 1.4 0.2 setosa
2 4.7 3.2 1.3 0.2 setosa
3 4.6 3.1 1.5 0.2 setosa
4 5.0 3.6 1.4 0.2 setosa

Train Test Split

Split your data into a training set and a testing set.

In [15]:
from sklearn.model_selection import train_test_split
In [42]:
X = iris.drop('species', axis =1)
y = iris['species']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.30)

Train a Model

Now its time to train a Support Vector Machine Classifier.

In [43]:
from sklearn.svm import SVC
In [44]:
svc_model = SVC()
In [45]:
X_train.shape
Out[45]:
(105, 4)
In [46]:
y.shape
Out[46]:
(150,)
In [47]:
svc_model.fit(X_train,y_train)
Out[47]:
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
  decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
  max_iter=-1, probability=False, random_state=None, shrinking=True,
  tol=0.001, verbose=False)

Model Evaluation

Now get predictions from the model and create a confusion matrix and a classification report.

In [48]:
predictions = svc_model.predict(X_test)
In [49]:
from sklearn.metrics import classification_report,confusion_matrix
In [50]:
print(confusion_matrix(y_test,predictions))
[[14  0  0]
 [ 0 16  0]
 [ 0  1 14]]
In [51]:
print(classification_report(y_test,predictions))
             precision    recall  f1-score   support

     setosa       1.00      1.00      1.00        14
 versicolor       0.94      1.00      0.97        16
  virginica       1.00      0.93      0.97        15

avg / total       0.98      0.98      0.98        45

You should have noticed this model was pretty good! Let's see if we can tune the parameters to try to get even better unlikely, and we probably would be satisfied with these results in real life because the data set is quite small.

Gridsearch Practice

Import GridsearchCV from SciKit Learn.

In [52]:
from sklearn.model_selection import GridSearchCV

Create a dictionary called param_grid and fill out some parameters for C and gamma.

In [53]:
param_grid = {'C': [0.1,1, 10, 100], 'gamma': [1,0.1,0.01,0.001]} 

Create a GridSearchCV object and fit it to the training data.

In [54]:
grid = GridSearchCV(SVC(),param_grid,refit=True,verbose=2)
grid.fit(X_train,y_train)
Fitting 3 folds for each of 16 candidates, totalling 48 fits
[CV] C=0.1, gamma=1 ..................................................
[CV] ................................... C=0.1, gamma=1, total=   0.0s
[CV] C=0.1, gamma=1 ..................................................
[CV] ................................... C=0.1, gamma=1, total=   0.0s
[CV] C=0.1, gamma=1 ..................................................
[CV] ................................... C=0.1, gamma=1, total=   0.0s
[CV] C=0.1, gamma=0.1 ................................................
[CV] ................................. C=0.1, gamma=0.1, total=   0.0s
[CV] C=0.1, gamma=0.1 ................................................
[CV] ................................. C=0.1, gamma=0.1, total=   0.0s
[CV] C=0.1, gamma=0.1 ................................................
[CV] ................................. C=0.1, gamma=0.1, total=   0.0s
[CV] C=0.1, gamma=0.01 ...............................................
[CV] ................................ C=0.1, gamma=0.01, total=   0.0s
[CV] C=0.1, gamma=0.01 ...............................................
[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s
[CV] ................................ C=0.1, gamma=0.01, total=   0.0s
[CV] C=0.1, gamma=0.01 ...............................................
[CV] ................................ C=0.1, gamma=0.01, total=   0.0s
[CV] C=0.1, gamma=0.001 ..............................................
[CV] ............................... C=0.1, gamma=0.001, total=   0.0s
[CV] C=0.1, gamma=0.001 ..............................................
[CV] ............................... C=0.1, gamma=0.001, total=   0.0s
[CV] C=0.1, gamma=0.001 ..............................................
[CV] ............................... C=0.1, gamma=0.001, total=   0.0s
[CV] C=1, gamma=1 ....................................................
[CV] ..................................... C=1, gamma=1, total=   0.0s
[CV] C=1, gamma=1 ....................................................
[CV] ..................................... C=1, gamma=1, total=   0.0s
[CV] C=1, gamma=1 ....................................................
[CV] ..................................... C=1, gamma=1, total=   0.0s
[CV] C=1, gamma=0.1 ..................................................
[CV] ................................... C=1, gamma=0.1, total=   0.0s
[CV] C=1, gamma=0.1 ..................................................
[CV] ................................... C=1, gamma=0.1, total=   0.0s
[CV] C=1, gamma=0.1 ..................................................
[CV] ................................... C=1, gamma=0.1, total=   0.0s
[CV] C=1, gamma=0.01 .................................................
[CV] .................................. C=1, gamma=0.01, total=   0.0s
[CV] C=1, gamma=0.01 .................................................
[CV] .................................. C=1, gamma=0.01, total=   0.0s
[CV] C=1, gamma=0.01 .................................................
[CV] .................................. C=1, gamma=0.01, total=   0.0s
[CV] C=1, gamma=0.001 ................................................
[CV] ................................. C=1, gamma=0.001, total=   0.0s
[CV] C=1, gamma=0.001 ................................................
[CV] ................................. C=1, gamma=0.001, total=   0.0s
[CV] C=1, gamma=0.001 ................................................
[CV] ................................. C=1, gamma=0.001, total=   0.0s
[CV] C=10, gamma=1 ...................................................
[CV] .................................... C=10, gamma=1, total=   0.0s
[CV] C=10, gamma=1 ...................................................
[CV] .................................... C=10, gamma=1, total=   0.0s
[CV] C=10, gamma=1 ...................................................
[CV] .................................... C=10, gamma=1, total=   0.0s
[CV] C=10, gamma=0.1 .................................................
[CV] .................................. C=10, gamma=0.1, total=   0.0s
[CV] C=10, gamma=0.1 .................................................
[CV] .................................. C=10, gamma=0.1, total=   0.0s
[CV] C=10, gamma=0.1 .................................................
[CV] .................................. C=10, gamma=0.1, total=   0.0s
[CV] C=10, gamma=0.01 ................................................
[CV] ................................. C=10, gamma=0.01, total=   0.0s
[CV] C=10, gamma=0.01 ................................................
[CV] ................................. C=10, gamma=0.01, total=   0.0s
[CV] C=10, gamma=0.01 ................................................
[CV] ................................. C=10, gamma=0.01, total=   0.0s
[CV] C=10, gamma=0.001 ...............................................
[CV] ................................ C=10, gamma=0.001, total=   0.0s
[CV] C=10, gamma=0.001 ...............................................
[CV] ................................ C=10, gamma=0.001, total=   0.0s
[CV] C=10, gamma=0.001 ...............................................
[CV] ................................ C=10, gamma=0.001, total=   0.0s
[CV] C=100, gamma=1 ..................................................
[CV] ................................... C=100, gamma=1, total=   0.0s
[CV] C=100, gamma=1 ..................................................
[CV] ................................... C=100, gamma=1, total=   0.0s
[CV] C=100, gamma=1 ..................................................
[CV] ................................... C=100, gamma=1, total=   0.0s
[CV] C=100, gamma=0.1 ................................................
[CV] ................................. C=100, gamma=0.1, total=   0.0s
[CV] C=100, gamma=0.1 ................................................
[CV] ................................. C=100, gamma=0.1, total=   0.0s
[CV] C=100, gamma=0.1 ................................................
[CV] ................................. C=100, gamma=0.1, total=   0.0s
[CV] C=100, gamma=0.01 ...............................................
[CV] ................................ C=100, gamma=0.01, total=   0.0s
[CV] C=100, gamma=0.01 ...............................................
[CV] ................................ C=100, gamma=0.01, total=   0.0s
[CV] C=100, gamma=0.01 ...............................................
[CV] ................................ C=100, gamma=0.01, total=   0.0s
[CV] C=100, gamma=0.001 ..............................................
[CV] ............................... C=100, gamma=0.001, total=   0.0s
[CV] C=100, gamma=0.001 ..............................................
[CV] ............................... C=100, gamma=0.001, total=   0.0s
[CV] C=100, gamma=0.001 ..............................................
[CV] ............................... C=100, gamma=0.001, total=   0.0s
[Parallel(n_jobs=1)]: Done  48 out of  48 | elapsed:    0.3s finished
Out[54]:
GridSearchCV(cv=None, error_score='raise',
       estimator=SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
  decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
  max_iter=-1, probability=False, random_state=None, shrinking=True,
  tol=0.001, verbose=False),
       fit_params=None, iid=True, n_jobs=1,
       param_grid={'C': [0.1, 1, 10, 100], 'gamma': [1, 0.1, 0.01, 0.001]},
       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',
       scoring=None, verbose=2)

Now taking that grid model and creating some predictions using the test set and create classification reports and confusion matrices for them.

In [56]:
grid_predictions = grid.predict(X_test)
In [58]:
print(confusion_matrix(y_test,grid_predictions))
[[14  0  0]
 [ 0 15  1]
 [ 0  0 15]]
In [59]:
print(classification_report(y_test,grid_predictions))
             precision    recall  f1-score   support

     setosa       1.00      1.00      1.00        14
 versicolor       1.00      0.94      0.97        16
  virginica       0.94      1.00      0.97        15

avg / total       0.98      0.98      0.98        45

We should have done about the same or exactly the same, this makes sense, there is basically just one point that is too noisey to grab, which makes sense, we don't want to have an overfit model that would be able to grab that.